Comparative Analysis of Black-Box Optimization Methods for Weather Intervention Design

TL;DR

The paper tackles the challenge of designing weather interventions under nonlinear, high-dimensional, and computationally expensive atmospheric dynamics by framing the problem as black-box optimization. It evaluates four methods—Bayesian optimization, random search, particle swarm optimization, and genetic algorithms—across two control paradigms (initial value intervention and model predictive control) using SCALE-RM in both a warm-bubble, idealized setting and a real-atmosphere scenario. Results show Bayesian optimization consistently yields the strongest precipitation reductions, especially in the higher-dimensional MPC context, and demonstrates robustness across random seeds. The work highlights BO as a promising tool for efficient weather-intervention design while acknowledging limitations in hyperparameter tuning and the feasibility of deploying such interventions in real-world atmospheric systems.

Abstract

As climate change increases the threat of weather-related disasters, research on weather control is gaining importance. The objective of weather control is to mitigate disaster risks by administering interventions with optimal timing, location, and intensity. However, the optimization process is highly challenging due to the vast scale and complexity of weather phenomena, which introduces two major challenges. First, obtaining accurate gradient information for optimization is difficult. In addition, numerical weather prediction (NWP) models demand enormous computational resources, necessitating parameter optimization with minimal function evaluations. To address these challenges, this study proposes a method for designing weather interventions based on black-box optimization, which enables efficient exploration without requiring gradient information. The proposed method is evaluated in two distinct control scenarios: one-shot initial value intervention and sequential intervention based on model predictive control. Furthermore, a comparative analysis is conducted among four representative black-box optimization methods in terms of total rainfall reduction. Experimental results show that Bayesian optimization achieves higher control effectiveness than the others, particularly in high-dimensional search spaces. These findings suggest that Bayesian optimization is a highly effective approach for weather intervention computation.

Figures (11)

• Figure 1: Accumulated precipitation over one hour without any control.
• Figure 2:
• Figure 3:
• Figure 4:
• Figure 5: